Project Summary/Abstract: The diagnosis of multiple sclerosis (MS) remains challenging due to its clinical heterogeneity and lengthy differential diagnosis. The incorrect assignment of a diagnosis of MS occurs in approximately 9% of newly evaluated patients and is associated with considerable clinically important, and avoidable, medical risk, morbidity, and healthcare costs. At the same time studies have demonstrated that many patients encounter a significant diagnostic delay prior to confirmation of a correct diagnosis of MS. In such patients early and accurate diagnosis of MS can result in prompt initiation of disease modifying therapy and consequent preventable disability. MS remains a clinical diagnosis and diagnostic criteria for MS are revised periodically, including most recently in 2017. Since implementation of the 2017 criteria, like all prior revisions, will continue to rely on subjective clinical and radiological assessments for its fulfillment, misdiagnosis will remain a risk. New objective, automated, and clinically applicable approaches to MS diagnosis are needed. Recent preliminary data from cross-sectional pilot studies in patients with established diagnoses have shown promise for three new radiographic and three new laboratory methods to differentiate MS from other disorders. The present study will evaluate these six methods for the first time in a prospective cohort of 125 patients undergoing an initial evaluation for MS at an academic MS subspecialty center. The specificity and sensitivity of each method will be compared to fulfillment of 2017 MS diagnostic criteria at the time of initial clinical evaluation. Using diagnostic thresholds developed from this analysis, a two year post-enrollment analysis will also be performed in participants who did not meet 2017 criteria initially but did so during the subsequent two year interval to determine if the study methods could have predicted a diagnosis of MS earlier in such patients. The use of a multimodal and machine-learning approach to evaluate the integration of each of these six new methods which represent different aspects of MS neuroinflammatory and neurodegenerative processes will also be performed during each analysis, and such a combination of radiographic and laboratory methodology may provide superior diagnostic accuracy compared to any given method alone. Planned collaborative career development, mentoring, and advising activities will facilitate acquisition of specific advanced quantitative and qualitative research skills necessary to develop and coordinate collection of data for this large prospective cohort study to rigorously evaluate new diagnostic methods for MS and incorporate machine learning analyses. Successful completion of this study will provide experience and skills necessary to move the field of MS diagnosis forward through a planned prospective multicenter NIH R01 funded study.